Don't ask to ask, just ask | |
PerlMonks |
Re^4: Better maps with Math::Geometry::Voronoi, and a Challenge for Math Monksby roboticus (Chancellor) |
on Jul 03, 2008 at 23:01 UTC ( [id://695491]=note: print w/replies, xml ) | Need Help?? |
BrowserUk:
Sorry, I should've been more clear. I wasn't referring to the vertices, but the input points to the algorithm. So the diagonal line going through two of the points are the bisector for the other two corners. I put this case into the algorithm (just a few minutes ago):
Yields the following lines:
That fourth entry is the diagonal I was talking about. It's clearly the bisector for points 0 and 3, but since it intersects at the same point as the first two, it's irrelevant. Looking at the edge list:
Other than the first edge, the rest are as you'd expect to see. I'm suspecting some odd comparison (or roundoff) in the C code to be generating that 0-length segment. Just for completeness, here's the list of vertices:
Looking at the results, I'm struck by a couple items:
...roboticus Update: I just tried the cross that you suggested, and the data showed *absolutely no* anomalies....odd! Update: Fixed HTML (missed closer for UL).
In Section
Meditations
|
|